Jesse's Notes

Zettelkasten

5 min read

Zettelkasten (German, “slip-box”) is a method for thinking and writing by building a web of small, linked notes rather than a hierarchy of long documents. It was popularised by the sociologist Niklas Luhmann, who kept a physical box of roughly 90,000 index cards and credited it with an improbable output — 70-plus books and 400-plus articles. The method’s claim is that the value lives not in any single note but in the connections between them: a well-tended slip-box becomes a thinking partner that surfaces links you’d forgotten you made.

The core principles

  • Atomic notes. One idea per note — small enough to be referenced on its own. Not “everything about caching” but a single claim. Atomicity is what makes notes recombinable; a note that holds three ideas can only be linked as a lump.
  • In your own words. You don’t transcribe a source; you rewrite the idea as you understand it. The rephrasing is the learning — it forces you to actually grasp the thing before you can store it.
  • Dense linking, with a reason. Every note links to related notes, and the link carries a sentence saying why they connect. The “because” is where the thinking happens; a bare link is just filing.
  • Emergent structure, not predefined folders. You don’t decide categories up front. Structure grows bottom-up out of the links. Luhmann used a handful of hub notes (cards pointing into a thread of the web) as entry points rather than a top-down taxonomy.
  • Stable addresses. Luhmann’s cards had IDs (21/3a, 21/3b…) so a new card could be slotted between two existing ones without reshuffling anything. In a digital tool the link graph does this job, so the explicit IDs matter much less.

The three note types

The widely-cited taxonomy comes from Sönke Ahrens’ How to Take Smart Notes, the book that carried Zettelkasten into the note-app era:

  1. Fleeting notes — quick captures, a thought before it escapes. Meant to be processed soon and then thrown away.
  2. Literature notes — what a specific source said, in your own words, kept with the reference. (Your reading-list/ is exactly this layer.)
  3. Permanent (evergreen) notes — the atomic, self-contained ideas that live in the slip-box forever and get linked into the web. (Your notes/ is this layer.)

The workflow is a refinement pipeline: fleeting → literature → permanent, with most material falling away and only the durable ideas earning a permanent note.

Why a slip-box compounds

The payoff is non-linear. Each new note can link to any existing note, so the number of possible connections grows faster than the number of notes. Past a threshold the box stops being a filing cabinet and starts being generative — when you sit down to write, the relevant ideas and their links are already assembled, because you did the connecting work at capture time rather than at deadline time.

The slip-box is not a collection of notes. Working with it is less about retrieving specific notes and more about being pointed to relevant facts and generating new connections. — paraphrasing Ahrens, How to Take Smart Notes

My read

This one barely needs adopting — my Quartz vault already is an informal Zettelkasten, and Quartz is built on exactly these ideas (atomic interlinked notes, an auto-generated backlink graph, hub index pages). Mapping it out:

  • In your own words → my standing habit when capturing a source: rephrase rather than transcribe, and add my own angle.
  • Literature vs. permanent notes → my reading-list/ (per-source) vs. notes/ (evergreen) split is the same distinction.
  • Dense linking → the wikilink conventions in this repo’s CLAUDE.md.
  • Hub notes as entry points → my index.md pages (the system-design hub, the Rails “mechanics” hub) are Luhmann’s index cards.
  • Compounding link graph → Quartz renders the backlinks for me.

So the honest framing is gaps, not adoption:

  1. Atomicity is uneven. Some notes — this section’s neighbour, the learning index — are grab-bags of several ideas plus loose links. A strict Zettelkasten would split those. But this is a trade-off, not a clear win: hub/index pages are meant to be collections, and over-atomising can shred things that read better together. I’d atomise concept notes, not hubs.
  2. Links don’t always carry a “because”. Plenty of my wikilinks just name the term and move on. Writing why two notes connect is the part I under-use, and it’s the part that does the actual thinking.

Where it sits among the other frameworks I’ve written up: these stack as input → knowledge → action, not competing choices.

  • PACER — how to absorb information while reading (retention).
  • Zettelkasten — how to store and connect the ideas worth keeping (this vault).
  • GTD — how to track what I need to do (the Daily Log).

Zettelkasten is the one describing the system I’m already running; PACER feeds it.

Further reading

Graph View